The resurrection of universal basic income (UBI) proposals in the developed world this year gained support from some prominent Australians. But while good in theory, it’s no panacea for the challenges of our modern economy.
UBI proposals centre on the idea that the government would pay a flat fee to every adult citizen, regardless of his or her engagement in skill-building activities or the paid labour market, as a partial or complete substitute for existing social security and welfare programs.
Of the schemes run in developing places like Kenya, Uganda, and India, some have been evaluated statistically, delivering some evidence of positive impacts on educational investments, entrepreneurship, and earnings.
In the developed world, Canada is trialling a UBI scheme. Finland also just rolled out a UBI trial, involving about 10,000 recipients for two years and costing about A$40 million. While Switzerland’s voters just rejected a UBI proposal via referendum, a similar proposal is presently looking like a goer for Utrecht in the Netherlands.
Here in Australia, it has been suggested the government might hand out somewhere between A$10,000 and A$25,000 a year to every man and woman.
Can we afford it?
There are two big questions to ask before taking a UBI proposal seriously, and the first is the most obvious one: where would the money come from to pay for it?
The present Australian welfare system (excluding the Medicare bill of A$25 billion) costs around A$170 billion per annum. Our GDP is around A$1.7 trillion per year, so this welfare bill is about 10% of annual GDP.
Giving A$20,000 to every Australian adult (19 million people) would cost approximately A$380 billion. That’s a little over twice the present total cost we pay for the above-mentioned programs.
Economics journalist Peter Martin has suggested that abolishing the tax-free threshold would pay for a UBI scheme. Working out whether the abolition of the tax-free threshold would fully fund a UBI is non-trivial, given the complexity of changing marginal tax rates as income rises and the need to estimate how many taxpayers fall into which tax brackets, but one thing is sure: the income tax bill of most if not all earners would have to rise in order to fund a UBI.
Let’s look at what would happen to someone earning A$80,000 per year if we were to implement a UBI, abolish the tax-free threshold, and leave the marginal income tax rate kink points unchanged.
This person presently pays about A$18,000 in income tax – made up of a marginal tax rate of 19% on dollars from the tax-free threshold of A$18,200 to A$37,000, and a marginal tax rate of 33% on dollars from A$37,000 to A$80,000.
Under a UBI scheme involving the abolition of the tax-free threshold, that same individual would receive the UBI (say, $20,000) but would then face a 19% tax rate on ALL of the first A$37,000 of income, and then a marginal tax rate of 33% on dollars from A$37,000 to A$80,000, yielding a total tax bill of A$21,220.
This person would be better off under the UBI in terms of take-home pay: instead of A$62,000, he would get A$79,000 in his bank account. But is the additional revenue increment from abolishing the tax-free threshold enough to pay for the UBI, if we spread it across all earners? I’ve yet to see a hard-headed answer to this question.
We may need to increase taxes elsewhere to pay for a UBI – possibly corporate taxes, land taxes, etc. – and most of these other taxes disproportionately impact richer people. Would legislation to fund a UBI scheme via increasing taxes on the rich get passed?
Some might suggest taking the money we presently spend on social security and welfare payments and converting it to a UBI. This would be enough to fund payments of about A$10,000 to each adult. But it would be a reverse-Robin-Hood policy: instead of spreading a fixed sum across our neediest citizens, we’d be spreading that sum across everyone, making the neediest worse off in order to send cash to our more well-heeled citizens.
Question 2: What is the perceived “broken” part of our present welfare system, and would UBI fix it?
Any targeted (e.g. means-tested) social support payment must be clawed back as income and/or wealth rises. While arguably an efficient way to get money to where it’s most needed, means-tested social security payments inevitably depress the incentive to earn more, at least in the section of the earning distribution where social security payments are paid out. This is talked about a lot, but I haven’t seen a reliable cost estimate for Australia of the disincentivising effects of the claw-back of social security and welfare payments.
The administration costs of the present system of targeted payments, also an argument given for moving to a UBI, are estimated at about A$3-4 billion. Whether this includes all much-pilloried costs of “churn” is an open question. And a UBI scheme would also have administration costs, which have not yet been estimated.
Another frequent theme describes the modern economy as turbulent, with more part-time, casual roles and more employment uncertainty than in the past. The implication is that the way our present system compensates people in insecure work is inadequate. Yet a social support model like a UBI that makes it even easier for people to be precariously attached to the workplace carries the implication that the workplace is not good for them.
In fact, work is socially and psychologically supportive for many people. Studies have found negative psychological impacts from job loss and retirement that seem driven by the social aspects of working. Do we want to isolate our most vulnerable citizens even more?
Arguably the biggest problem in our current tax-and-transfer system is that the rich, and their organisations (like big companies), do not get taxed enough and/or benefit from special provisions (for example in regard to superannuation).
Instead of changing our present targeted welfare system into a diffuse scattergun money-for-all scheme, we could instead courageously tackle the worst part of the problem first.
What else is relevant?
Some worry that a UBI scheme would further depress the incentive to work. As I’ve stated elsewhere, I doubt that the drive to win in terms of labour market success is going to go away anytime soon.
Others have also found small impacts on work incentives from cash transfer programs, though this evidence is mainly drawn from developing countries.
In principle, two types of work incentives may be affected. People receiving unconditional handouts every year may feel less pressure to get and keep a job. Secondly, if the UBI were funded by the abolition of the tax-free threshold and/or increases to income tax rates, then people would be more strongly penalised for working additional hours and might hence work less.
Would people receiving unconditional handouts feel freer to explore their creative sides? To explore entrepreneurial ideas? To engage in more meaningful work than they presently do? These are all mentioned as possible benefits of a developed-country UBI, but we really don’t know whether they would materialise, nor do we really know how to measure them.
We do know that people adapt to new reference points (including income reference points), and there is good reason to expect that at least some of a long-term UBI would be soaked up in higher prices – particularly for goods that the poor buy most.
Watch and learn
A proposal to throw money at people, while wrapping that proposal in the flags of “equality” and “basic rights”, can be argued to be the lazy man’s face-saving response to the complex, entrenched problems of poverty. The poor arguably lack access and/or skills as much as or more than they lack money.
What’s more, the present Australian social security and welfare system can be viewed as a UBI scheme with exceptions for people who don’t need it. Some changes to the system that do not involve wholesale overhauls could address many of the problems discussed above.
My advice for Australia? Watch the policy experiments in Europe keenly. But don’t assume for one minute that universal basic income is a magic bullet. Compared to our current system, it is expensive, inefficient, and potentially regressive.